Challenges in Mechanistic and Empirical Modelling of Stormwater: Review and Perspectives

Abstract

Much research has been conducted to quantify stormwater runoff and quality using both mechanistic (i.e. process-based) and empirical (i.e. data-driven) techniques. Mechanistic models generally include the mathematical representation of relevant physico-chemical processes to generate storm runoff quantity and quality. Empirical approaches analyse available data for potential response and predictor variables to trace the interactions of major processes and develop data-driven explanatory and/or predictive relationships. This paper reviews major, mostly unresolved, challenges with both mechanistic and empirical modelling of stormwater, sheds light on the scientific gaps with conventional practices, and offers important perspectives by taking the highly urbanized Miami River Basin of Florida as an analytical example. Appreciating the varying levels of process complexity in different urban river basins, we discuss the relative applicability of mechanistic and empirical methods for robust predictions of stormwater quantity and quality. RÉSUMÉBeaucoup de recherches ont été menées pour quantifier les eaux pluviales générées par les tempêtes, et leur qualité. Elles utilisent à la fois des techniques mécaniques (basées sur les processus) et empiriques (déduites des données). Les modèles mécanistes comprennent généralement la représentation mathématique des processus physico-chimiques pertinentes pour générer la quantité et la qualité des eaux de ruissellement en condition de tempête. Les approches empiriques analysent les données disponibles pour que potentiellement les variables de réponse et les variables prédictives retracent les interactions entre les processus majeurs, et développent des relations explicatives et/ou prédictives pilotées par les données. Ce document passe en revue les principaux défis, pour la plupart non résolus, avec la modélisation à la fois mécaniste et empirique des eaux pluviales; il met en lumière les lacunes scientifiques avec les pratiques conventionnelles et offre d'importantes perspectives en prenant comme exemple d'analyse le très urbanisé bassin de la rivière Miami, en Floride. En appréciant le niveau de complexité variable des processus dans des différents bassins hydrographiques urbains, nous avons discuté de l'applicabilité relative des méthodes mécanistes et empiriques pour des prédictions robustes de la quantité et de la qualité des eaux pluviales.

Smart management of combined sewer overflows: From an ancient technology to artificial intelligence

“…The mechanistic stormwater and CSO models comprise the mathematical representation of the relevant real‐case situations and generate results based on the principles of physics and chemistry (Al‐Amin & Abdul‐Aziz, 2013). These models require moderate or extensive input data and most of the time the input data comes with several uncertainties.…”

Irrigation Canal Network Flow Analysis by a Hydraulic Model

“…Traditionally, hydraulic simulation models of an irrigation area are used for water resources planning and management (Al‐amin and Abdul‐aziz, ; Shrestha et al ., ) and for determining proper irrigation and drainage strategies by calculating water requirements (Bayat et al ., ; Karamouz et al ., ). In South Korea, HOMWRS (Hydrological Operation Model for Water Resources System), a reservoir operation model, has been frequently used for managing irrigation systems.…”

New Frontiers of Hydrology: Soil, Water, and Vegetation Monitoring and Modelling

“…The second paper, by Fernando Nardi, Chiara Biscarini, Silvia Di Francesco, Piergiorgio Manciola and Lucio Ubertini, presents the investigations conducted for floodplain identification in the Tiber river basin in central Italy by means of a DEM‐based hydrogeomorphic method, which was compared to the use of standard flood maps for preliminary mapping of river valleys at the global scale (Nardi et al ., ). The third paper, by Shams Al‐Amin and Omar I. Abdul‐Aziz, discusses both mechanistic numerical and empirical data‐driven methods for stormwater quantitative estimation, offering a detailed review while also underlining the current scientific gaps and practical issues of projects dealing with flood water management for highly urbanized areas under intense hydrologic forcing, such as the presented case study of the Miami river basin of Florida (Al‐Amin and Abdul‐Aziz, ). A further paper, the fourth in the supplement, by Gabriella Petaccia and Elisabetta Natale, introduces ORSADEM, a one‐dimensional (1‐D) shallow‐water hydraulic model able to represent shocks and discontinuities generated by a dam break (Petaccia and Natale, ).…”